Question about Rigid Connectors

[Marloes vdH]

Dear all,

I am working with the oks models for my Master Thesis project, and am applying a gait cycle through the rigid connectors. Imagionary rigid bodies are created to create the Rigid connector, a rigid cylindrical joint, with which two of the 6 Degrees of Freedom can be described.

Now I have the following simplified situation:

I fixed my tibia in all DoF and apply a compressive force using the rigid connector for the Internal-External rotation DoF (and translation along this z axis).
The rigid connector is defined between the tibia and the imagionary body TFTO.
The joint origion of the rigid connector is the same as the center of mass of the TFTO AND the tibia.
The simulation results in a downwards translation of my femur, thus of my femur on my tibial cartilage.

What I don't understand is HOW this rigid connector between the tibia and TFTO can result in forces on my femur?
Just to be clear, I want this to happen, but I do not understand how my femur is coupled to this rigid connector.

It is not clear to me how this works, even after reading the FEBio theory manual. I tried to mimic this rigid connector in a simplified model I created myself with two rigid bodies and one imagionary body. However, I only get the imagionary body to translate and nothing to happen with my 'femur rigid body'

Is there someone you can clarify how these imagionary bodies in the rigid connectors work and how they result in forces in my femur?

Kind regards,
Marloes

[Snehal Chokhandre]

Hi Marloes,

The three cylindrical joints create a linkage system between the tibia and femur. When you apply compressive force at cylindrical joint between the tibia and TFTO (applying force on tibia), you apply an equal and opposite force on the femur through the linkages. So, if you removed one of the rigid cylindrical joint definitions in the FeBio file, the linkage will break and femur will not move.

I hope that helps.

Best,
Snehal

[Marloes vdH]

Hi Shenal,

Thanks for your kind reply, this helps a lot!
So if you apply the force through the cylindrical joint, at what point is the force applied? At the origin of the cylindrical joint? Because if this would be the case, the location of the origin could lead to a moment as a result of this compressive force. For example, when the origin is shifted medial, could you get a Varus/Valgus moment in the knee? If the cylindrical joint only enables to describe the translational and rotational DoF along the zaxis, could a compressive force even lead to a moment in other DoF?

I tried to mimic this in the Gen1 model and saw no such thing happen. It seems the force is applied to the whole rigid joint and cannot lead to moments resulting from a compression/tension. Is this the case?

So this linkage system is created by the three cylindrical joints, but how is this linkage created mathematically?

Kind regards,
Marloes

[Marloes vdH]

Hi Shenal,

If i understand the FeBio theory manual correctly, the force is applied to the origin, but translation is only possible along the z-axis, thus the location has no influence anymore since no valgus torque will result from this compression. Is this correct?

The reason I ask this is because when applying compression to my oks models(001&003), there is much more lateral contact pressure than medial. I did change the contact conditions to be able to see the contact pressure in my medial tibial cartilage but stil, the lateral tibial cartilage had a higher contact pressure.

When looking into literature, often there is more medial contact pressure than lateral. It would seem that the mechanical axis of the joint should have an angle in stead of staying (0,0,1) but I am afraid that this would influence the External-Internal rotations too much.

Do the Rigid connectors have the right joint coordinate axis? If not, would it be reasonable to change the axis of the Ext-Int cylindrical joint? Or

Kind regards,
Marloes

[Snehal Chokhandre]

Hi Marloes,

Sorry about the delay.
Yes, In cylindrical connectors, translations and rotations are limited along/around the axis.
For the oks models you are using, the IE axis is along the z- axis for convenience ( anatomical landmarks were picked from 3D geometries we created for the specimens, to define the coordinate systems). in the future versions of the models, the coordinate systems will be defined based on anatomical landmarks probed during the experimentation. You can find the details here,
https://simtk.org/plugins/moinmoin/open ... ystems.pdf
Once, the axes are oriented as they were during experimentation, we hope to see a better correspondence to the experimental data and hopefully to the literature.

Thanks,
Snehal

[Marloes vdH]

Hi Shenal,

That explains a lot, thank you!
When I look into the link you send, it is stated that the linkage is a result of the kinematic chain equation. But if I look into this kinematic chain equation, it would seem that a force in one cylindrical joint could induce a translation in another. Is that correct?

So these anatomical landmarks probed during experimentation can be obtained through simVITRO. Are these coordinate systems already obtained and documented for some of the oks models, or is this yet to be done?

Kind regards,
Marloes

[David W]

Dear all,

I am also working on the OKS model for my Msc Thesis,with the goal of eventually incorporating musculoskeletal modelling output as FEBio model input.

To do so I have been running some test simulation for the last couple of weeks to get familiar with the model and especially the function of the Rigid Connectors. I am trying to apply a compressive force of -100N axially and afterwards have the knee flex while under this load.

With the goal of eventually incorporating the flexion-extension rotation and compressive force extracted from my MSM models using a load curve.
I want the secondary kinematics of the knee to arise from the soft tissue interactions as a result of the flexion-extension rotation and the compressive force.

For my test simulation I have first applied a displacement step of -0.15 mm in the Z direction by applying a -0.15 mm translation to the lmgLnk2-Tibia joint to make sure the femur makes contact with the menisci (step1). Afterwards when I apply the -100N to the lmgLnk2-Tibia joint (step2) I keep getting inverted elements resulting in a Negative Jacobian Error.

I have also tried applying a Rigid force of -100N to the Femur (step2), I am able to get this to work by constraining the Femur in all DoFs except for the Z-displacement in the force step.

For the rotation step (step 3) I want to apply a rotation of 45 degrees to the Femur-lmgLnk1 joint while under the -100N load. In this step the Femur is not constrained using a Rigid Constrained as I believe the Rigid Connectors should constrain the model. Using this tactic I am not able to get the model to work.

Would someone be able to explain me why I am unable to apply a load and subsequently a rotation to the cylindrical joints? Constraining the model further for mee seems to defeat the purpose of the rigid connectors.Thanks in Advance

David

[Snehal Chokhandre]

Hi David,

It is a little challenging to figure out convergence issues in a complex model like this, they may be resulting from over/under constrained rigid bodies among other things.

I would recommend you use one of the generation 2 models found here - https://simtk.org/plugins/datashare/?group_id=485# (specifically in OKS_models, 'final model' folders) to troubleshoot further. There are a total of 8 models customized for a 90 degree passive flexion. You can update them for your application if you choose to use one of them. There are several differences / improvements over the generation 1 model that you are using. For example, these models also have quadriceps tendon- patella - patellar ligament and, medial and lateral patellofemoral ligaments representations. This may allow a little more stability.

You can find the details of the model development here - https://simtk.org/svn/openknee/app/Knee ... ations.pdf (main document describing the whole process of model development) and https://simtk.org/svn/openknee/app/Knee ... ations.pdf ( this documents details any changes in strategy that were employed once the actual model was developed ( for specimen oks003 ) and carried on to the rest of the models).

You can compare any of the generation 2 model files to the generation 1 file you are using to see the differences in settings, tolerances, contact definitions etc. to get some more insight into what might be causing the convergence issues. These models were run using Febio 2.9.

I hope this helps.

Thanks,
Snehal